Note: Descriptions are shown in the official language in which they were submitted.
WO 90/11287 20466,49 PCT/US90/01526
-1-
MATRIX METALLOPROTEINASE INHIBITOR PEPTIDES
Field of the Invention
This invention relates to isolated proteins or
peptides useful for inhibition of matrix metalloproteinases.
Specifically, this invention relates to a novel protein,
isolated from conditioned media of cultured human tumor
cells, which binds with high affinity to matrix
metalloproteinase enzymes and analogs thereof. The natural
protein is defined by a novel amino acid sequence including
specific positions of cysteine residues. This invention
further relates to a novel means of purifying matrix
metalloproteinase inhibitors using metalloproteinase affinity
chromatography.
Backaround of the Imaention
The collagenase family of enzymes are a group of
neutral metalloproteinases, also known as matrix
matalloproteinases, which are secreted in the zymogen form
and degrade both the collagenous and noncollagenous
components of the extracellular matrix. All require a metal
ion (calcium and/or zinc) for hydrolytic activity, and all
are secreted in the latent pre-enzyme form. Members of this
collagenase gene family include: the interstitial
collagenases, which degrade collagen types I, II and III and
have been characterized with respect to substrate specificity
and requirements for activation (Stricklin, G.P., Jeffrey,
J.J., Rosewit, W.T., and Eisen, A.Z., 1983, Biochemistry 22,
61-68; Goldberg, G.I., Wilhelm, S., Kronberger, A., Bauer,
E.A., Grant, G.A., and Eisen, A.Z., 1986, J. Biol. Chem. 261,
6600-6605; Hasty, K.A., Jeffrey, J.J., Hibbs, M.S., and
Welgus, H.G., 1987, J. Biol. Chem. 262, 10048-1052; Fields,
G.B., Van Wart, H.E., and Birkedal-Hansen, H., 1987, J. Biol.
Chem. 262, 6221-6226; Grant, G.A., Eisen, A.Z., Marmer, B.L.
Rosweit, W.T., and Goldberg, G.I., 1987, J. Biol. Chem. 262,
5886-5889); stromelysin, which degrades proteoglycans,
glycoproteins, and the non-helical portions of collagenous
molecules (Wilhelm, S.M., Collierm, I.E., Kronberger, A.,
WO 90/11287 2046649 PCT/US90/01526
-2-
Eisen, A.Z., Marmer, B.L., Grant, G.A., Bauer, E., and
Goldberg, G.I., 1987, Proc. Natl. Acad. Sci. U.S.A. 84, 6725-
6729; Whitman, S.E., Murphy, G., Angel, P., Rahmsforf, H.-
J., SMith, B.J., Lyons, A., Harris, T.J.T., Reynolds. J.J.,
Herrlich, P. and Docherty, A.J.P., 1986, Biochem. J. 240,
913-916); and type IV collagenase, which degrades pepsin-
resistant triple-helical type IV collagen and interstitial
collagens (gelatin). Type IV collagenase has been identified
in human tumor cells (Liotta, L.A., Kleinerman, J.,
Catanzaro, P., and Rynbrandt, D., 1977, J. Natl. Cancer Inst.
58, 1427-1439; Turpeenniemi-Hujanen, T., and Tryggvason, K.,
1982, Int. J. Cancer 30p, 669-673; Liotta, L.A., Abe, S.,
Gehron-Robey, P., and Martin, G.R., 1979, Proc. Natl. Acad.
Sci. U.S.A. 76 2268-2272; Liotta, L.A., Tryggvasson, K.,
Garbisa, S., Hart, I., Foltz, C.M., and Shafie, S., 1980,
Nature (London) 284, 67-68; Collier, I.E., Wilhelm, S.M.,
Eisen, A.Z., Marmer, B.L., Grant, G.A., Seltzer, J.L.,
Kronberger, A., He., C., Bauer, E.A., and Goldberg, G.I.,
1988, J. Biol. Chem. 263, 6579-6587), endothelial cells
(Kalebic, T., Barbisa, S., Glaser, B., and Liotta, L.A.,
1983, Science 221, 281-283), bone (Murphy, G., McAlpine,
C.G., Poll, C.T., and Reynolds, J.J., 1985, Biochem. Biophys.
Acta 831, 49-58), fibroblasts (Collier, I.E., Wilhelm, S.M.,
Eisen, A.Z., Marmer, B.L., Grant, G.A., Seltzer, J.L.,
Kronberger, A., He., C., Bauer, E.A., and Goldberg, G.I.,
1988, J. Biol. Chem 263, 6579-6587), polymorphonuclear
leukocytes (Uitto, V.J., Schwartz D., and Veis, A., 1980,
Eur. J. Biochem. 105, 409-417) and macrophages (Garbidsa, S.,
Ballin, M., Daga-Giordini, D., Fastelli, G., Naturale, M.,
Negro, A., Semenzato, G., and Liotta, L.A., 1986, J. Biol.
Chem. 261, 2369-2375). This enzyme is a neutral
metalloproteinase of 68 to 72 kilodaltons which is secreted
in zymogen form (Liotta, L.A., Abe, S., Gehron-Robey, P., and
Martin, G.R., 1979, Proc. Natl. Acad. Sci. U.S.A. 76, 2268-
2272; Liotta, L.A., Tryggvassin, K., Garbisa, S., Gehron-
Robey, P., and Abe, S., 1981, Biochemistry 20, 100-104; Salo,
T., Liotta, L.A., and Tryggvsasson, K., 1983, J. Biol. Chem.
WO 90/11287 2046649 ' PCT/US90/01526
-3-
258, 3058-3063). In addition, several other members of this
collagenase gene family have been described recently,
including a second type of stromelysin (stromelysin-2), a 92
kilodalton form of type IV collagenase, and Putative Uterine
Metalloproteinase (PIUMP)-1, a low molecular weight uterine
collagenase (Wilhelm, S.M., Collier, I.E., Marmer, B.L.,
Eisen, A.Z., Grant, G.A., and Goldberg, G.I., 1989, J. Biol.
Chem. 264, 17213-17221; Woessner, J. F. and Talpin, C. J.,
1988, J. Biol. Chem. 263, 16918-16925).
The matrix metalloproteinases are thought to play an
important role in disease processes characterized by the
inappropriate destruction of the extracellular matrix.. The
diseases include inflammatory processes such as rheuinatoid
arthritis and other autoimmune disorders, tumor cell invasion
and metastasis formation, local sequelae of myocardial
anoxia, and corneal ulceration (Okada et al., 1986, J. Biol.
Chem. 261, 14245-14255; Harris et al., 1984, Collagen Relat.
4, 493-512; Werb et: al., 1977, New Engl. J. Med. 296,
1017-1023; Liotta et al., 1980, Nature (London) 284, 67-68;
Kalebic et al., 1983, Science 221, 281-283). Many tissues
contain natural inhibitors of the matrix metalloproteinases.
In some cases, this inhibitory activity is derived from the
antiproteases in plasma, particularly a2-macroglobulin and Ri-
anticollagenase. a2-Macroglobulin is a high molecular weight
(725,000 Da) inhibitor present in serum. It is thought to
account for 95% of the collagenolytic inhibitory activity
present in serum. Because of its large size, it is normally
unable to pass the vascular permeability barrier. Under
conditions of extreme inflammation in which there is
increased capillary permeability, a2-macroglobulin may enter
the tissue compartments and play a role in the regulation of
matrix metalloproteir,Lases. The mechanism of inhibition of
the matrix metalloprot:einases by a2-macroglobulin has not been
directly studied. However, it is thought to be similar to
the mechanism whereby a2-macroglobulin causes inhibition of
WO 90/11287 2046649 PCT/US90/01526
-4-
other proteases. Hence, the mechanism is not believed to be
unique for the matrix metalloproteinases.
(3i-Anticollagenase is approximately 40,000 daltons in
size. It accounts for approximately 5% of the
metalloproteinase inhibiting activity of serum. This
inhibitor is thought to pass the vascular permeability
barrier and be widely distributed in the tissue compartments.
(3l-Anticollagenase may be related to another group of natural
inhibitors of the metalloproteinases referred to as TIMPs,
tissue inhibitors of inetalloproteinases. The down-regulation
of metalloproteinase collagenolysis and proteolysis may occur
through TIMPs.
The prototype TIMP, TIMP-1, is a glycoprotein with an
apparent molecular size of 28.5 kDa which forms a complex of
1:1 stoichiometry with activated interstitial collagenase,
stromelysin, and the 92 kDa type IV collagenase (Welgus and
Stricklin, 1983, J. Biol. Chem. 253, 12259-12264: Welgus et
al., 1985a, Collagen Rel. Res. 5, 167-179; Wilhelm et al.,
1989, J. Biol. Chem. 264, 17213-17221; European patent
189,784). The gene coding for TIMP-1 has been cloned,
sequenced and mapped to the X-chromosome (Carmichael et al.,
1986, Proc. Natl. Sci. USA 83, 2407-2411; Docherty et al.,
1985, Nature (London) 318, 66-69; Mullins et al., 1988,
Genomics 3, 187-194; Mahtani et al., 1988, Genomics 2, 294-
301). The secreted protein has 184 amino acids and six
intramolecular disulfide bonds. Reduction and alkylation of
TIMP-1 abolishes all inhibitory activity. The same cells
which produce interstitial collagenase are capable of
synthesizing and secreting TIMP-1 (Welgus et al., 1985b, J.
Clin. Invest. 76, 219-224; Herron et al., 1986, J. Biol.
Chem. 261, 2814-2818). Thus, the net collagenolytic activity
for these cell types is the result of the balance between
activated enzyme levels and TIMP-1 levels. Studies have
shown an inverse correlation between TIMP-1 levels and the
invasive potential of murine and human tumor cells. Down-
modulation of TIMP-1 mRNA levels by use of TIMP-1 antisense
RNA resulted in conversion of previously nontumorigenic,
WO 90/11287 204U60 P(,'I'/US90/01526
-5-
noninvasive Swiss 3T3 cells to tumorigenic cells with
invasive properties in vitro and metastatic potential in vivo
(Khokha et al., 1989, Science 243, 947-950).
Another class of biologically active collagenase
inhibitors is composed of low molecular weight (>10,000
daltons) cationic proteins isolated from cartilage, aorta and
teeth, but which are poorly characterized.
Recently several new members of the matrix
metalloproteinase family have been identified, with various
substrate specificities. These include stromelysin
(homologue of rat transin), type IV collagenase (70 kDa
gelatinase) and a 92 kDa gelatinase. While also identified
in normal cell types, the over-expression of these enzymes
has been linked to malignant conversion and the metastatic
phenotype in a number of systems. Thus, there is a need to
understand the molecular basis of the regulation of these
metalloproteinases aind to find inhibitors which can be
exploited for diagnostic and therapeutic purposes.
SUMMARY OF THE INVENTION
It is the object of this invention to provide means
of purifying natural inhibitors of metalloproteinases.
It is a further object of this invention to provide
matrix metalloproteinase inhibitors and derivatives thereof.
The inhibitors may be obtained from natural sources, may be
produced by synthetic: means such as the Merrifield peptide
synthesis process, or by genetically engineered organisms or
cell lines. The inhibitors of the invention may be used to
treat disease conditions which result from activity of matrix
metalloproteinases. Furthermore, since metalloproteinase
activity is essential to the implantation of the zygote,
these inhibitors are useful as contraceptives.
The present invention relates to novel
metalloproteinase inhibitors distinct from previous
inhibitors mentioned above. Described herein are the
isolation and sequencing of a novel protein, now designated
TIMP-2 (initially called CSC-21K, so named for its amino
terminal amino acid sequence and apparent molecular weight on
WO 90/11287 2046649 PCT/US90/01526
-6-
gel electrophoresis). CSC-21K forms a 1:1 complex with type
IV procollagenase and type IV collagenase and inhibits
activated type IV collagenase. Binding of CSC-21K to
activated type IV collagenase results in inhibition of its
collagenolytic activity. This inhibitor can be isolated
using affinity chromatography on purified matrix
metalloproteinase attached to a solid phase. Amino acid
sequence analysis of CSC-21K reveals significant homology to
TIMP-1 indicating that CSC-21K is the first novel additional
member of the family of TIMP-like proteins.
Thus, a preferred embodiment is a protein of
approximately 21,600 daltons which binds to matrix
metalloproteinases and can be isolated using affinity
chromatography on solid phase purified metalloproteinases.
The amino acid sequence of this isolated protein shows that
it is a new gene product not previously discovered and has
areas of sequence homology with the known natural tissue
inhibitor of inetalloproteinases (TIMP-1). The protein of the
preferred embodiment of this inhibitor is characterized by
the amino acid sequence shown in Figure 5, below.
DESCRIPTION OF THE FIGURES
Figure 1. Anion exchange chromatography of the
complex of collagenase type IV and inhibitor isolated from
human melanoma cell (A2058) conditioned media and eluted from
gelatin affinity chromatography. 15 pg of gelatin-affinity
purified material were applied to the anion exchange resin.
The column was eluted with a linear gradient of NaCl (-- ).
Material from the single major peak eluting at 0.18 M NaCl
was rechromatographed on the reverse phase column (insert).
Material from peaks A and B were sequenced directly (see Fig.
2B).
Figure 2. A. 15% polyacrylamide-SDS gel
electrophoresis of CSC-21K and CSC-21K-type IV collagenase
complex. Lane A. 2 pg of CSC-21K (peak A) material
following reverse phase HPLC purification. Lane B. 2}lg of
CSC-21K-type IV collagenase complex isolated by gelatin-
WO 90/11287 2 0 4 6 6 4 9 pLT/US90/01526
-7-
Sepharose affinity chromatography. Gel was run at 25
milliamps constant current, using a Laemmli sample buffer
system and sample buffer containing (3-mercaptoethanol.
Samples were heated at 95 C for 2 min prior to
electrophoresis. B. Amino terminal amino acid sequence of
reverse phase HPLC peaks. The complex obtained following
gelatin-affinity and anion exchange chromatography was
further purified into components by reverse phase HPLC,. The
materials obtained in peaks A and B (Fig. 1, insert) were
sequenced directly.
Figure 3. CSC-21K inhibition of activated type IV
collagenase/gelatinase activity. A. Dose relationship of
purified CSC-21 (upper curve with, lower curve without
reduction and alkylation) inhibition of purified, p-APMA
activated type IV collagenase. CSC-21K is termed TIMP-2 and
presented as a mole/mole basis. The substrate is native type
IV collagen. B. Dose relationship of purified CSC-21K
(termed TIMP-2) inhibition of purified p-APMA activated type
IV collagenase. The substrate is gelatin.
Figure 4. CSC--21K protein sequence data obtained from
the amino terminus and following digestions with
endoproteinases Lys-C, Arg-C and Asp-N. Peptide sequences
obtained following digests were aligned by overlaps
(underlined regions) as shown. The entire sequence of CSC-
21K is encompassed by these overlapping peptides. The origin
of each of the peptides is identified in the lower half of
the figure.
Figure 5. Complete sequence for CSC-21K (TIMP-2)
derived from direct amino acid sequencing and homology to
human TIMP-1. Computerized homology searches using the
BIONET system were applied to the sequence obtained following
digestions with endop:roteinases Lys-C, Arg-C and Asp-N. The
results of these homology searches are shown.
Figure 6. cDNA sequence and deduced protein sequence
of clones pSS15 and pSS18 which encode portions of CSC-21K.
Figure 7. Nucleotide sequence and prediction amino
acid sequence of a complete human TIMP-2 cDNA. The cDNA
WO 90/11287 PCI'/US90/01526
2046649
-8-
insert of clone pSS38 was sequenced in both directions using
dideoxy-methodology. The predicted amino acid sequence is
shown under the DNA sequence. The putative polyadenylation
signal is underlined.
Figure 8. Comparison of TIMP-2 deduced amino acid
sequence and direct amino acid sequencing of CSC-21K protein
(see Figure 5, above). CSC-21K primary structure was
determined directly using a Porton Instruments 2020 gas phase
protein sequenator and phenyl hydantoin derivative
identification ori a Beckman System Gold HPLC unit equipped
with a 0.46 X 25 cm Beckman ODS column. Comparison shows 96%
identity of these sequences. Asterisks identify changes in
the sequence identified by DNA sequencing of the complete
TIMP-2 cDNA.
Figure 9. Homology comparison of TIMP-2 and TIMP-1
at the amino acid (A) and nucleotide levels (B). A. Deduced
amino acid sequences of TIMP-2 and TIMP-1 were compared using
a Pustell Scorinci Matrix. The analysis was performed using
a cutoff value of' 66% homology, and an 8 amino acid overlap.
The line denotes regions in which the homology exceeds the
average value of 66% homology between these two protein. B.
Comparison of the: nucleotide sequences of TIMP-2 and TIMP-1.
Analysis was per!'ormed using a Pustell Scoring Matrix, with
a hash value of 4 and a window of 30. The line indicates
regions of ident:Lty. The analysis if performed for TIMP-1
vs. TIMP-1 or TIMP-2 vs. TIMP-2 gives a solid line on the
diagonal, indicating complete identity. This demonstrates
that TIMP-2 is a unique gene product distinct from TIMP-1.
Figure 10. Northern blot analysis of TIMP-2 mRNA
expression in cultured cell lines. Total cytoplasmic and
oligo-dT selectedRNA as isolated from cells as described in
the text. After transfer to Nytran* filter RNA was hybridized
with 32P-labelled probe specific for TIMP-2. The resulting
autoradiographs and shown. A). Oligo-dT selected RNA (1ug)
from A2058 human melanoma cells. B). Total cytoplasmic RNA
(54g) from WI-38 human embryonic fibroblasts (lane 1) and HT-
1080 human fibrosarcoma cells (lane 2).
* Trade Mark
L ~.
WO "/11287 2 0 16- 64 9 PC.'I'/US90/01526
-9-
Figure 11. Northern blot analysis of total
cytoplasmic RNA isolated from A2058 melanoma cells following
48 hour treatment with either 12-tetradecanoylphorbol
13-acetate (10 ng/mL, lane B) or transforming growth factor
R1 (5ng/mL, lane C). These are compared with basal levels in
untreated A2058 cells (lane A). Equal amounts of RNA (5 pg)
were loaded and the e'thidium bromide stained gel is shown as
a control (insert).
Figure 12. Northern blot analysis of human colorectal
tumors and adjacent normal mucosa. RNA (5 ug) of each sample
was electrophoresed aiid transferred as described in the text.
Lanes Tl, T2 and T3 contain RNA from the invasive colorectal
tumors. Lanes Nl, N2 nd N3 contain RNA from the corresponding
adjacent normal mucosa.
DETAILED DESCRIPTION OF THE INVENTION
The present invention relates to an inhibitor of
matrix metalloproteinases for which the determination of the
complete primary structure shows that this protein is a
second member of the TIMP family, TIMP-2, as recently
reported (Stetler-Stevenson et al., 1989, J. Biol. Chem. 264,
17374-17378). TIMP--2 is a 21 kDa protein which selectively
forms a complex with the latent proenzyme form of the 72 kDa
type IV collagenase (Stetler-Stevenson et al., 1989, J. Biol.
Chem. 264, 17374-17378; Goldberg et al., 1989, Proc. Natl.
Acad. Sci. USA 86, 8207-8211). The secreted protein has 192
amino acid residues azzd is not glycosylated. TIMP-2 shows an
overall 71% homology to TIMP-1 at the amino acid sequence
level. The positior.L of the twelve cysteine residues are
conserved with respect to those present in TIMP-1, as are
three of the four tryptophan residues. TIMP-2 inhibits the
type IV collagenolytic activity and the gelatinolytic
activity associated with the 72 kDa enzyme. Inhibition
studies demonstrated ithat complete enzyme inhibition occurred
at 1:1 molar ratio of TIMP-2 to activated 72 kDa type IV
collagenase enzyme (Stetler-Stevenson et al., 1989, J. Biol.
Chem. 264, 17374-172178). Thus unlike TIMP-1, TIMP-2 is
WO 90/11287 2 0 4 6 6 4 9 PCI'/US90/01526
-10-
capable of binding to both the latent and activated forms of
type IV collagenase. Cell culture studies using cell lines
that produce a variety of collagenase family enzymes, as well
as both TIMP-1 and TIMP-2, suggest that TIMP-2 preferentially
interacts with the 72 kDa type IV collagenase (Stetler-
Stevenson et al., 1989, J. Biol. Chem. 264, 17374-17378;
Goldberg et al., 1989, Proc. Natl. Acad. Sci. USA 86,
8207-8211). Thus, like interstitial collagenase activity
which is the balance of activated enzyme and TIMP-1, the net
72 kDa type IV collagenase activity may depend upon the
balance between the levels of activated enzyme and TIMP-2.
Analogs of the natural inhibitor of the invention can
be made by preparing peptides and proteins having cysteines
at the same intervals as the cysteines in the natural
inhibitor. Other amino acids may vary from the pattern of
the natural inhibitor so long as the cysteine is located at
the appropriate intervals. At least two appropriately spaced
cysteines must be present in the peptide to ensure inhibitory
activity by virtue of a disulfide bridge formation.
While the preferred protein contains the sequence of
Figure 5, peptides having amino acids identical with the
sequence of Figure 5 in at least 50% of the sequence
positions are within the scope of the invention having useful
inhibitor of metalloproteinase activity provided the
cysteines are retained in the desired relative positions.
Peptide fragments derived from the natural CSC-21
molecule were used as immunogens. In the case of synthetic
peptide fragments of the protein to be used as immunogens or
antigens for antibodies specific for CSC-21K, one skilled in
the art understands that a unique amino acid sequence in
terms of recognition by an antibody binding site consists of
a sequence of from four to six amino acids which is not known
to exist in another protein in the environment for which the
antibody is to be used (e.g., human biological specimens).
Further, a unique nucleotide sequence in this context refers
to the nucleotide sequence encoding a unique amino acid
sequence as defined above and, therefore, consists of from
WO 90/11287 PCT/US90/01526
2Q-46649
four to six codons (12 to 18 nucleotides) needed to encode
four to six amino acids. Antibodies to such unique protein
fragments can be used to detect the natural inhibitor in
serum, tissue, and other natural sources.
Particularly preferred peptides are those having at
least 2 cysteines. An amino acid sequence containing the
sequence CSCSPVHPQQAFCNA derived from the amino terminal of
the molecule and segments containing the amino acid sequences
SLNHRYQQGCECKITRC]? and MIPCYISSPDECLWTD appear particularly
active. The ant:Lgenic and functional utility of peptides
derived from CSC-21K is not limited to these peptides but can
include the whole protein, natural or synthetically derived.
Example 1
Purification of CSC-21K (TIMP-2)
Human A2058 melanoma cells were grown to 80%
confluence in Du:Lbecco's modified Eagle's medium with 10%
fetal bovine serum. The medium was then replaced with serum-
free Dulbecco's modified Eagle's medium, and the culture
continued for an additional 24 hours. Approximately 60 L of
human melanoma cell (A2058) serum-free conditioned medium was
concentrated to 300 mL using an Amicon* YM 30 ultrafiltration
membrane. This concentrated conditioned medium was applied
to two 1.0 x 10 cm gelatin-Sepharose* (Sigma Chemical Co.)
affinity columns in series, equilibrated with 0.05 M Tris
HC1, 0.5 M NaCl, 0.005 M CaC121 0.02% Brij 35, pH 7.6 buffer.
The columns were then washed with equilibration buffer before
eluting with 10.0% DMSO in equilibration buffer. The eluate
was concentrated and exchanges into 0.05 M Tris HC1, 0.15 M
NaCl, 0.005 M CaC121 0.02% Brij 35, pH 7.6 using as Amicon* YM
30 membrane. The samples were stored at -80 C. Samples for
anion exchange chromatography were dialyzed into 0.01 M Tris
HC1, pH 7.5, witlh 20% ethylene glycol. A 15 ug sample was
injected into a Dionex A1400 HPLC system equipped with a 0.4
x 5.0 cm Dionex Propac* anion exchange column. This column
was eluted with a.linear gradient of zero to 0.4 M NaCl. The
material under the single major peak was collected and an
aliquot was applied to a 0.46 x 10 cm RP300 column (Pierce
Trade Mark
;
0
WO 90/11287 PCT/US90/01526
2046649
-12-
Chemical Co.). This column was eluted as previously
described (U.S.Patent No. 5,270,447, Krutzsch et al.).
Alternatively, the
complex as obtained from gelatin-Sepharose chromatography and
stored at -80 C can be applied directly to the RP300 column
system.
CSC-21K was isolated as a complex with human type IV
collagenase by gelatin affinity chromatography of human
melanoma cell (A2058) conditioned media. Anion exchange
chromatography of'the material obtained from gelatin-affinity
chromatography i=esulted in a single species eluting at
approximately 0.18 M NaCl (Figure 1). Reverse phase HPLC
analysis of material eluted from the ion exchange
chromatography showed that this material contained two
components (Figure 1, insert). The material obtained from
the gelatin-affinity chromatography step is thus an
intermolecular complex as seen on anion exchange
chromatography, and is not a simple copurification of two
species on gelati_n-affinity chromatography. NaDodSO4-PAGE of
the complex obtained from the gelatin-affinity chromatography
also showed two components, Figure 2A. The higher molecular
weight material has an M, of 70,000. It was identified as
type IV procollagenase by immunoblotting and amino terminal
sequencing (vide infra). The lower molecular weight material
has an apparent Mr of 18,000 which increased to 21,000 upon
reduction. Direct reverse phase HPLC analysis of the complex
obtained from ge:Latin affinity chromatography resulted in the
separation of two peaks identical to those in the insert of
Figure 1. The inaterial obtained from each of these peaks,
designated peak.A for the material with the shorter retention
time and peak B for the material of longer retention time,
was subjected to amino acid analysis and direct amino acid
sequencing. Peak A material gave a unique amino-terminal
amino acid sequence shown in Figure 2B. This material is
referred to as CSC-21K. Peak B material gave an amino-
terminal sequence identical to latent type IV collagenase
(i.e., type IV procollagenase), Figure 2B.
WO 90/11287 2 0 46,649' PCT/US90/01526
-13-
Example 2
Enzyme digestions, amino acid seauencina, and
amino acid composition analyses
HPLC purified CSC-21K was reduced and alkylated as
described. 15 ug of reduced and alkylated CSC-21K was
incubated with 5 ug of endoproteinase Lys-C, 5 pg of
endoproteinase Arg-C, or 2}lg of endoproteinase Asp-N in 0.1
M NH4HCO3 buffer overnight at 37 C. The digests were then
separated by reverse phase HPLC on the RP300 column into
component peaks which were collected and sequenced
individually. Amino acid sequence analysis was carried out
on HPLC-purified fractions on a Porton Instruments 2020 Gas
Phase Protein Sequenator using standard program 39. PTH
amino acid identification was carried out on a Beckman System
Gold HPLC unit equipped with a 0.46 x 25 Beckman ODS column
and eluted using a modified sodium acetate/THF/acetonitrile
separation method.
Amino acid composition analyses were performed
following vapor phase: hydrolysis for 18 h using 6N HC1, 0.1%
phenol at 120 C. The hydrolysate was derivatized using the
PITC method (PicoTag system, Waters) and analyzed in the same
HPLC unit as above using a modified triethylamine/ammonium
acetate/acetonitrile elution method.
Amino acid composition analyses of the complex eluted
from the gelatin-aff'inity chromatography, and CSC-21K are
compared in Table 1. The amino acid composition of CSC-21K
is significantly different from other collagenase inhibitors
and is distinguisheci by an unusual Leu/Ile ratio. This
feature was used to evaluate the stoichiometry of the complex
as isolated by gelatin affinity chromatography. Based on the
experimentally determined molar amino acid composition of
CSC-21 (7 Leu, 18 Ile; Table 1, which is in agreement with
the direct amino acid sequence from overlapping peptides,
Figure 4) and the deduced composition of type IV
procollagenase (39 Leu/25 Ile), it was calculated that the
theoretical Leu/Ile ratio of a 1:1 molar complex would be 46
WO 90/11287 2046649 PC17/US90/01526
-14-
Leu/42 Ile or 1.10. This is in excellent agreement with the
ratio value of 1.03 that was determined from the amino acid
composition analysis of the CSC-21K-type IV procollagenase
complex. Thus human melanoma cells, which are known to
secrete several metalloproteinases, also secrete a protein,
CSC-21K, which specifically binds to the latent form of type
IV collagenase and forms a complex with 1:1 molar
stoichiometry.
WO 90/11287 PCr/US90/01526
2046649
-15-
Table. 1. Amino acid compositions of CSC-21K (TIMP-2) and
the CSC-21K+Collagenase IV Complex
CSC-=21K CSC-21K Complex
---=-------------------------------
AMINO ACID RESIDUES
RESIDUE Picamoles(l) MOLE(2) Picamoles(l)
ASP/ASN 141 18 149
GLU/GLN 166 22 127
SER 107 14 78
HIS 32 4 25
GLY 222 29 123
ARG 56 7 61
THR 59 8 86
ALA 108 14 92
PRO 93 12 92
TYR 60 8 64
VAL 86 11 67
MET 11 1 28
CYS 33 4 12
ILE 137 18 83
LEU 57 7 81
PHE 50 7 86
LYS 127 17 107
TOTAL 1545 201 1361
LEU/ILE RATIO 0.42 0.39 1.02
(i)Data obtained from direct amino acid composition analysis
of reduced ar.Ld alkylated CSC-21K, or enzyme inhibitor
complex as isolated by gelatin-Sepharose' chromatography,
as described above.
(Z)Molar amino acid composition calculated from amino acid
composition data assuming 7 phenylalanine residues per mole
of CSC-21K.
The complete primary structure of human CSC-21K, determined
by sequence ar-alysis of overlapping peptides obtained
following endoproteinase Lys-C, endoproteinase Arg-C and
endoproteinase Asp-N digestions, is shown in Figure 4. The
amino acid composition of CSC-21K determined from this
sequence data concurs with that obtained by direct analysis
* Trade Mark
A.;
WO 90/11287 PCT/US90/01526
2046649
-16-
of purified CSC-21K, Table 1. The molecular weight of CSC-
21K calculated from the primary sequence is 21,600 daltons,
which is in good agreement with the gel electrophoresis data.
Computer searches for homology were performed on the BIONET
Protein Data Base: (which accesses the NBRF-PIR and SWISS-PROT
protein sequence data banks). Computerized homology searches
were applied to the entire peptide sequence. This provided
the basis for alignment of the CSC-21K structure with that of
human TIMP, Figure 5. CSC-21K shares significant homology
with TIMP. ThE:re is 41.0% amino acid identify and 29%
conservative substitutions in a 191 amino acid overlap. The
positions of the twelve cysteine residues are conserved and
the positions of three of four tryptophan residues are also
conserved. The conservation of the relative positions of
these residues supports their functional or structural roles
in both proteins.
Example 3
Collaaenolvtic and Qelatinolvtic inhibition.
<~y Type IV collagenase assays were performed as
previously described (U.S.Patent No. 5,270,447,
Krutzsch et al.).
Gelatinase assays were performed by adaptation of this method
utilizing heat denatured rat skin collagen (NEN/Dupont). The
CSC-21-collagenase IV proenzyme complex was activated by a 1
h preincubation with 1 mM p-aminophenylmercuric acetate (p-
APMA). Subsequeritly, purified CSC-21K was added prior to the
assay of collageinase IV activity.
As isolated following gelatin-affinity chromatography,
the complex between CSC-21K and type IV procollagenase
possessed no collagenolytic activity. Following activation
with the organomercurial compound p-amino-phenylmercuric
acetate (p-APMA), the maximum achievable type IV
collagenolytic activity obtained was 7.12 }ig type IV collagen
degraded/h/pg enzyme complex. The maximum gelatinolytic
activity obtained following organomercurial activation was
26.4 pg/h/p.g of enzyme complex. Reduction of disulfide bonds
i y
WO 90/11287 PCT/US90/01526
20 466 49
-17-
destroys the complex formed between TIMP and interstitial
collagenase and also destroys the complex between CSC-21K and
type IV procollagenase (Figure 3A). The addition of purified
native CSC-21K, but not reduced and alkylated CSC-21K, to
this p-APMA activated complex resulted in proportional
inhibition of both collagenolytic and gelatinolytic
activities (Figure 3 A and B). Extrapolation of this data
demonstrated that the binding of CSC-21K to the activated
enzyme occurs in a stoichiometric manner that is consistent
with the 1:1 molar ratio determined for the complex isolated
by gelatin-Sepha.rose' chromatography. These results
demonstrate that. CSC-21K which has not been exposed to
organomercurial compounds is capable of binding to and
inhibiting activated type IV collagenase suggest, but they
also suggest that p-APMA activation of type IV collagenase
may be accompanied by the organomercurial-mediated
inactivation of CSC-21K.
These data show that while CSC-21K shares scattered
homology with TIMP-1, particularly with respect to conserved
positions of the cysteine residues, all of the CSC-21K
peptides are distinctly different from sequences of the known
TIMP-1. Thus, the peptides of the invention are encoded by
a gene different from that which encodes TIMP. This
demonstrates that CSC-21K is the product of a separate gene.
Synthetic peptides were prepared using the sequence
from the amino terminal portion of the CSC-21K molecule.
These were coupled to bovine serum albumin for use in
generating anti-peptide antibodies by standard methods. The
antibodies were affinity purified using solid phase peptide-
affinity chromatography as previously described (Stetler-
Stevenson et al., 1989, J. Biol. Chem. 264:1353-1356). These
antibodies are reactive on standard western and immunoblots.
The isolated, purified CSC-21K, recombinant CSC-21K,
and analogs can be used therapeutically in this diseases
characterized by the uncontrolled activity of matrix
metalloproteinases. Such diseases include arthritis,
diabetes, cancer, ulcers of mucosa and epithelial tissues,
* Trade Mark
~1. ~~
WO 90/11287 2046649 PCT/US90/01526
-18-
autoimmune mediated inflammation, lung injury, granulomatous
diseases. A particularly useful application may be in the
treatment of myocardial infarctions since matrix proteolysis
including destruction of the myocardial basement membrane is
a harmful process in this affliction. Other therapeutic
benefit might also be obtained in diseases with basement
membrane destruction such as lupus, autoimmune neural
disorders, myocyte destruction such as myodystrophies,
myocardial infarct and glomerulopathies. CSC-21K could also
be used as a potential birth control agent by preventing
embryo/placental attachment or invasion.
Example 4
CloninQ of Human CSC-21K.
Human A2058 melanoma cells were grown to confluence
was run over an oligo dT column to selectively isolate
messenger RNA species. This mRNA preparation was then used
to prepare a cDNA library using the LambdaGem-4 vector and
standard methodology. 1ug of purified mRNA was used to
prepare double stranded cDNA using a commercially available
cDNA synthesis kit (Amersham). This cDNA was methylated
using EcoRI methylase (Promega), linked to EcoRI linkers
(Promega), restricted with EcoRI and ligated to EcoRI
digested Lambda-GEM-4 (Promega). The ligations were packaged
(Gigapack Gold, Stratagene) and the optimal reactions were
pooled to give 1.5 X 106 recombinants. 7.5 X 105 recombinants
were screened using oligonucleotide 27-40. Oligonucleotide
27-40, a 45-mer, with the sequence: 5'-
GAGAAGGAGGTGGACTCTGGCAATGACATCTATGGCAACAACATC-3',
corresponding to the reverse translation of residues 27
through 40 of the previously sequenced TIMP-2 protein.
Oligonucleotide 27-40 was synthesized on a Biosearch 8700 DNA
synthesizer by means of R-cyanoethyl phosphoramidite
chemistry, and was labelled using y-[32P]-ATP (Amersham) and
T4 kinase (Bethesda Research Laboratories). From the total
of 750,000 plaques screened, 239 positives were identified.
Of these positives, initially eight clones were further
WO 90/11287 204- 'U6'~~9 PGT/US90/01526
-19-
characterized following SpeI digestion of the parent
LambdaGem-4 clones, religation of the SpeI digests and
ampicillin resistance: selection of the transformants. These
eight clones were cross-screened following southern blot
hybridization with four additional synthetic oligonucleotides
that were also based on the protein sequence data for CSC-
21K. Only two clones reacted positively with all four
additional synthetic oligonucleotide probes. These clones
are designated pSS15 and pSS18. The larger of these two
clones is pSS15 which is a pGEM-1 vector containing a 2.1 Kb
insert. This clone contains an internal HindIiI restriction
site located approximately 1.2 Kb from the 5' end of the
clone. This insert can be released from the pGEM-1 vector by
dual endonuclease restriction with EcoRI and XbaI.
Both clones pSS15 and pSS18 were subcloned into M13
and sequenced using the dideoxy method. The pSS15 clone was
subcloned using the two HindIIl fragments. The partial cDNA
sequences obtained and the deduced amino acid sequences are
shown in Figure 6. The amino acid sequence obtained is
identical, within the limits of experimental error, to that
obtained for a portion of CSC-21K shown in Figure 5. These
results demonstrate that these clones referred to as pSS15
and pSS18 encode the protein CSC-21K. It is obvious that due
to differences in codon preferences between species that
clones from other species could encode functional CSC-21K
protein but with a different nucleotide sequence. Thus, one
base change per codon. of the CSC-21K cDNA, resulting i.n 33%
change in the overall nucleotide sequence, may still result
in a cDNA which would encode a functional CSC-21K protein.
Thus, the existence of this clone for human CSC-21K is a
reduction to practice of isolating the cDNA encoding this
protein from other species.
A deposit of CSC-21K cDNA (pSS15) has been made at the
American Type Culture Collection (ATCC), 12301 Parklawn
Drive, Rockville, Maryland, 20852, U.S.A., on August 11, 1989
under the accession number 40,644. The deposit shall be
viably maintained, replacing if it became nonviable, for a
WO 90/11287 20 4664 9 PGT/US90/01526
-20-
period of 30 years from the date of the deposit, or for 5
years from the last date of request for a sample of the
deposit, whichever is longer, and made available to the
public without restriction in accordance with the provisions
of the law. The Commissioner of Patents and Trademarks, upon
request, shall have access to the deposit.
In subsequent experiments, two addition clones were
isolated and the nucleotide sequence of the cDNA insert in
the clone pSS38, the longest, is presented in Figure 7. The
insert contains 730 bp, excluding the poly(AT) tail and
encodes the mature TIMP-2 protein of 194 amino acids. The
130-nucleotide long 3' untranslated region contains the
putative polyadenylation signal 30 bases upstream from the
3' end of the RNA.
Comparison of the amino acid sequence of TIMP-2
deduced from the cDNA clone with that determined by direct
amino acid sequencing of overlapping endoproteinase derived
peptide fragments shows excellent agreement. The original
sequence contained only 192 amino acids. The previously
unidentified residues correspond to the glycyl residue at
positions 92 and the prolyl residue at the carboxyl terminus.
Other changes are noted in Figure 8. The homology of TIMP-
2 with TIMP-2 at the predicted amin acid sequence level is
37.6% identity and 65.6% overall homology. Pustell Matrix
analysis of the homology distribution between these two
predicted protein sequences using a cutoff value of 66% and
an 8 amino acid overlap, demonstrates that there are two
areas in which the homology falls below this average value.
TIMP-2 shows a distinct preference for binding to the latent
form of the 72 kDa type IV collagenase in the presence of
both other latent metalloproteinases and TIMP-1 (Stetler-
Stevenson et al., 1989, J. Biol. Chem. 264, 17374-17378;
Goldberg et al., 1989, Proc. Natl. Acad. Sci. USA 86,
8207-8211). However, both forms of TIMP will inhibit
activated type IV collagenase. Thus regions of amino acid
sequence that are highly conserved between these proteins,
such as those that exceed the overall homology value of 66%,
WO 90/11287 204 6649 PCT/US90/01526
-21-
may be responsible for the known shared functions of these
proteins, inhibition of the activated collagenase family
enzymes which are unique for individual TIMP molecules.
Thus, the regions of low homology between residue 20 to 45
and the carboxyl terniinus of TIMP-2, may be responsible for
the binding of TIMP-2 to the latent form of the 72 kDa type
IV collagenase.
Example 5
Applications of the TIMP-2 cDNA.
Comparison of the cDNA sequence of human TIMP-2 with
that of human TIMP-1 shows little homology compared to that
seen at the amino acid level, Figure 9b. This result implies
that these genes diveirged early in the evolution of this gene
family. The lack of homology at the cDNA level may also
explain why TIMP-2 mRNA transcripts are not detected in
northern blot analyses using TIMP-1 probes and also why
screening cDNA libraries with TIMP-1 probes fails to yield
TIMP-2 clones.
Northern blot analysis of oligo-dT selected mRNA
isolated from various cells have been carried out using the
TIMP-2 cDNA. HT-1080 human fibrosarcoma cells, WI-38 human
embryonic lung fibroblasts, and A2058 human melanoma cells
were grown to 80% confluence in Dulbecco's modified Eagle
medium (DMEM, GIBCO). The medium was then replaced with DMEM
supplemented with 0.5% ITS+ (Collaborative Research Inc.) and
25 }ig/mL Gentamycin. The medium was changed after 4 hours
and culture continued for 20 hours prior to the addition of
10 ng/mL TPA (Sigma Chemical Co.) or 5 ng/mL TGF-01 (R & D Systems).
Total cytoplasmic RNA was isolated from cell lines as
described (Gough 1988, Anal. Biochem. 173, 93-95). mRNA was
isolated using the FAST-TRACK mRNA isolation kit
(Invitrogen). Tissues mRNA was isolated from frozen tissue
fragments. Tissue fragments were obtained from three partial
colectomy specimens at the time of surgery, from Dr. Barry
Schmuckler, Washingtori Hospital Center, Washington, D.C.. The
pathologic diagnosis of all three cases was invasive
adenocarcinoma. Tissue samples were also obtained from
WO 90/11287 PCT/US90/01526
2046649
-22-
adjacent, uninvolved mucosa. Frozen tissue was pulverized in
liquid N2 using a mortar and pestle. The tissue powder was
then dissolved in 4 M guanidine isothiocyanate, 3 M sodium
acetate, 0.84% P-nnercaptoethanol, pH 6Ø Total cytoplasmic
RNA was isolated by pelleting through 5.7 M cesium chloride,
3 M sodium acetate, pH 6Ø Aliquots of RNA were applied to
formaldehyde/1% w/v agarose gels and electrophoresed before
transfer onto Nytran' filters (Schleicher & Schuell). The RNA
was UV-crosslinked to the filter and hybridized to the insert
from clone pSS38. The pSS38 cDNA probe was labelled with a-
[32P]-dCTP using a random primer labelling kit (Bethesda
Research Laboratories).
Northern blot analysis of the A2058 human melanoma
cell line revealed two specific mRNA species with approximate
sizes of 3.5 and 0.9 kb (Figure 10a). These mRNA species
were also detected in RNA isolated from human WI-38
fibroblasts, with very low levels of the 0.9 kb species
detectable in eiquivalent amounts of RNA from HT-1080
fibrosarcoma cells (Figure 10b). The origin of these two
specific transcripts remains to be determined, however the
size difference is too large to be easily accounted for by
differences in 3' polyadenylation. It is possible that
alternative 5' uritranslated regions could account for the
different transcript sizes, as has been demonstrated for
insulin-like growth factor II mRNA's.
Treatment of A2058 cells with 12-0-
tetradecanoylphorbol 13-acetate (TPA) (10 ng/ml) for 48 hours
failed to significantly modulate TIMP-2 transcript levels
(Figure 11). This is in contrast to mRNA level for the 72
kDa type IV collagenase, which was down regulated in response
to TPA. Interstitial collagenase mRNA is rapidly induced
following TPA treatment of A2058 melanoma cells and
fibroblast cells lines (Chin et al., 1985, J. Biol. Chem.
260, 12367-12376; Werb et al., 1986, J. Cell Biol. 102, 697-
702; Frisch et a.l., 1987, Proc. Natl. Acad. Sci. USA 84,
2600-2604), as is TIMP-1 (Edwards et al., 1985, Mol. Cell.
Biol. 5, 3280-32813; Murphy et al., 1985, J. Biol. Chem. 260,
Trade Mark
WO 90/11287 20 4 6649 PCT/US90/01526
-23-
3079-3083; Welgus et al., 1985b, J. Clin. Invest. 76, 219-
224). Treatment of A2058 melanoma cells with transforming
growth factor-beta 1(TGF-01) for 48 hours resulted in a
clearly detectable decrease in TIMP-2 mRNA levels (Figure
11). The 3.5 and 0.9 kb transcripts showed equal decreases
in steady-state level, and there was no indication of
differential expression. TGF-01 has been demonstrated to
increase TIMP-1 mRNA levels in human gingival fibroblasts
(Overall et al., 1989, J. Biol. Chem. 264, 1860-1869). It
has been demonstrated previously that TGF-01 induced the 72
kDa type IV collagenase mRNA and protein levels, as well as
enzyme activation. In the presence of the other qrowth
factors, TGF-01 also has a selective reciprocal effect on
interstitial collagenase and TIMP-1 expression (Edwards et
al., 1987, EMBO J. 6, 1899-1904). TGF-01 selectively
represses the induction of interstitial collagenase, but
interacts synergistically to super-induce TIMP-1. These data
demonstrate that TIMP=-1 and TIMP-2 respond differently to TPA
treatment and oppositely to TGF-01 treatment. Furthermore,
TGF-01 has a reciprocal effect on TIMP-2 and the 72 kDa type
IV collagenase transcript levels in human melanoma cells.
Thus it is clear that the transcriptional regulation of TIMP-
2 is independent of TIMP-1.
Finally, northern blot analysis of tissue from three
primary, human colorectal tumors and adjacent normal mucosa
was performed using the pSS38 TIMP-2 probe and is presented
in Figure 12. The matched samples showed no detectable
change in TIMP-2 naRNA transcript levels between the
colorectal tumor samples and the adjacent normal mucosa.
Previous studies have shown that indeed human colorectal
tumor tissue contains elevated type IV collagenase mRNA
transcripts. These dlata suggest that in the primary tumor
cell population the ratio of TIMP-2 to 72 kDa type IV
collagenase is altered in favor of the enzyme species by
differential transcription. However, due to primary tumor
cell heterogeneity and the possible inclusion of normal cell
WO 90/11287 2046649 PCT/US90/01526
-24-
populations in the invasive tumor samples, this observation
may not accurately reflect the invasive, metastatic cell
subpopulations. Examination of metastatic lesions will allow
a better understanding of the role of these proteins in tumor
cell invasion.
Utility of the CSC-21K cDNA Clones
The isolated human cDNA clones encoding the
metalloproteinase inhibitor protein, TIMP-2, have wide
utility in diagnostics. Pathologic conditions including
neoplasia, inflammatory diseases, cardiovascular disease,
central nervous system disorders, diabetes and abnormalities
of growth and development may be accompanied or causally
related to abnormal levels of metalloproteinase inhibitor
protein, the subject of invention. All of these processes
may involve abnormal accumulation or loss of extracellular
matrix proteins. In particular, many of these disease states
exhibit abnormal basement membranes. Since control of
basement membrane breakdown may be regulated by inhibitors of
metalloproteinase action, the inhibitor protein of the
subject invention may play a key role in determining the
steady-state levels of basement membranes. The present cDNA
clones encoding the inhibitor protein CSC-21K can be used in
northern blotting analysis to measure the mRNA levels of the
inhibitor in RNA samples isolated from tissue samples or
cultured cells, as described in Example 5 (above). In some
cases, elevated CSC-21K mRNA levels detected in this fashion
may reflect pathologic states leading to increased basement
membrane accumulation, such as diabetes mellitus. In other
cases, loss of the inhibitor protein may be important, such
as in neoplasia and central nervous system disorders
involving the basement membrane surrounding nerves. In
addition to hybridization of the isolated cDNA clone (whole
or in part) with isolated RNA or DNA, in situ hybridization
using tissue or cell samples can be readily conducted using
methods well known in the art. For this and other purposes,
the clone can be labeled with radioactive markers for
WO 90/11287 2046649 PCT/US90/01526
-25-
detection by using suitable enzymes and radioactive
precursors.
The inhibitor protein CSC-21K is a suppressor of
neoplastic invasion, and as such, is a tumor suppressor gene.
Therefore, homozygous loss, allelic loss, or mutational
inactivation of the gene regulatory region may suppress
expression of the inhibitor and favor the development of
cancer. All of these genetic defects can be detected using
the isolated cDNA clone in standard Southern blotting
analysis, with or without prior polymerase chain reaction
amplification of sample DNA sequences, with methods and
appropriate restriction enzymes well known in the art.
Extraction of DNA and measurement of such genetic defects may
be useful in the diagnosis of cancer and the detection of
individuals with hereditary defects predisposing to the
development of cancer.
The isolated cDNA clones may be useful in genetic
therapy. Diseases associated with loss or down regulation
of expression of the subject inhibitor protein could be
treated with the cDNA clone in a suitable expression vector,
allowing augmented synthesis of the CSC-21K protein.
Transfection of the cDNA clone for CSC-21K, in an expression
vector with a suitable promoter, into a cell deficient in
CSC-21K production would result in the increased production
of CSC-21K and correction of the abnormal phenotype.
Alternatively, antisense constructs, using the same
expression vector but containing the reverse orientation of
the CSC-21K cDNA insert, could be used to suppress the
overproduction of metalloproteinase inhibitor protein. This
could be useful in disorders of abnormal regulation or
inappropriately high production of the CSC-21K inhibitor
protein. The systerri and methods of preparation of such
genetic reagents is known in the art, but requires the
specific isolated nucleotide sequence of the invention. It
is obvious that the cDNA clone of the present invention could
be spliced next to a gene encoding any other protein to
produce a hybrid protein. This methodology could be used to
W0 90/11287 2 0 4 6 6 4 9 pCF/US90/01526
-26-
produce a hybrid protein with enhanced inhibitory activity or
tumor-seeking behavior.
The cDNA clone, or any other DNA segment of this
invention which encodes the amino acid sequence of the TIMP-
2 protein (according to the universal genetic code) is
necessary and highly useful for the recombinant production
of inhibitor protein CSC-21K (TIMP-2). Recombinant CSC-21K
can be made in any suitable expression system, either
prokaryotic or eukaryotic. A significant advantage in the
production of the present invention is that the protein is
not glycosylated, and does not require post-translational
modification for functional activity. Thus, recombinant
proteins made in bacterial expression systems can be
functionally active directly as obtained from the culture
medium. It is obvious that the recombinant inhibitor protein
can be linked to suitable carrier proteins, marker proteins
or other compounds which stabilize or potentiate its
activity. The recombinant protein whole, or in part, can be
used as an antigen or treatment agent to inhibit
metalloproteinase action.
The cDNA clone of the subject invention encodes a
novel metalloproteinase inhibitor CSC-21K. The gene itself
is novel and distinct encodes a novel metalloproteinase
inhibitor CSC-21K. The gene itself is novel and distinct
from all prior art reporting cDNA clones encoding proteinase
inhibitors. In fact, hybridization of any or all of the cDNA
clones for proteinase inhibitors existing in the prior art,
under either stringent conditions or conditions of reduced
stringency, fails to detect the gene of the present
invention, whole or in part. For that reason, the gene of
the present invention has never before been detected. The
affinity purification of this protein, and the identification
of the novel amino acid sequence of this invention was a
totally original approach, not described in the prior art
which led to the isolation of the gene of the present
invention.
WO 90/11287 2 04S 64 9 PCT/US90/01526
-27-
It is well known in the field of protein chemistry
that the functional properties of a protein do not depend on
the identity of 100% of the amino acid residues which
comprise the protein. Individual amino acid residues can be
substituted which have the same charge or hydrophobicity and
achieve the same function. Furthermore, other amino acids in
a protein molecule which specifically function to determine
the protein structure can substituted with residues which
differ in charge or hydrophobicity without diminishirig the
overall biological activity of the parent protein molecule.
More generally, the degree of evolutionary similarity of the
amino acid sequences of two structurally and functionally
related polypeptide:; is determined by the method of
quantitative analysis defined by the sequence alignment and
comparison algorithnis described by Pearson and Lipman
(Pearson, W. R. & Lipman, D. J., 1988, Proc. Nat. Acad. Sci.
U. S. A. 85:2444-48). This quantitative comparison
contemplates not orily precise homology of amino acid
sequences, but also substitutions of one residue for another
which are known to occur frequently in families of
evolutionarily related proteins sharing a conserved function.
Thus, in the present case, this invention also relates to an
isolated polypeptide which inhibits a matrix
metalloproteinase and. which has an amino acid sequence that
differs in at least orie position from the sequence defined in
Figure 7 and yet has greater similarity to the amino acid
sequence of Figure 7, or to a unique portion thereof, than to
the amino acid sequence of any other polypeptide.
When matrix metalloproteinase inhibitors of the
invention are used in the treatment of inappropriate
angiogenesis, arthritis, tumor growth, invasion and
metastasis, and granulomatous inflammatory conditions such
as sarcoidosis and other pathological conditions, it is
possible to estimate the amount of enzyme produced and the
amount of peptide inhibitor required to inhibit greater than
90% of the active enzyme. For use in treating any disease
condition, the therapeutic dose of the inhibitory peptide
WO 90/11287 2 Q 4 6f f 4 9 P(,'I'/US90/01526
-28-
falls within an acceptable pharmacologic range of 1-250
mg/kg/da, with a more preferred dosage being 25-100 mg/kg/d.
The dosage for a given patient will depend on the amount of
enzyme produced in the patient, the condition and size of the
patient. The inhibitors may be given as infusions or by any
means which provides ready transmission into the circulation.
Lyophilized powders may be "snorted". Preparations for
buccal or sublingual administration may also be given. For
respiratory tract involvement, the peptides may be
administered by inhalation. Aerosols are particularly useful
for this purpose. For conditions of the eye, the peptides
may be administered as eye drops.
The isolated CSC-21 proteins, natural or recombinant,
or active peptides derived therefrom can be administered
intravenously, orally, intrauterine, by inhalation or topical
application. For example, topical application can be
prepared using a suitable carrier for treatment of basal cell
carcinomas or melanomas of the skin or for the treatment of
corneal ulceration.
The complete CSC-21 protein or CSC-21 peptides can be
produced by purification from natural sources, by synthetic
peptide chemistry methods or by recombinant DNA technology.
In the latter case, suitable cDNA clones for CSC-21 in a
suitable expression vector can be used to produce peptides
with CSC-21 activity.
CSC-21 peptides and antibodies to CSC-21 are also
useful in diagnosis of diseases characterized by abnormal
balances of matrix metalloproteinase and associated
inhibitor. Purified CSC-21 may be used by virtue of its
ability to bind metalloproteinases as a means to purify and
or detect metalloproteinases from any natural source.
Suitable immunoassays for CSC-21 could include anti-CSC-21
antibodies, reference CSC-21 antigen and solid or solution
phase reactions. Purified CSC-21 or peptide domains of CSC-
21 can be tagged with suitable enzymatic, fluorescent or
radioactive labels by means well known in the art.
WO 90/11287 PCT/US90/01526
2046649
-29-
Peptides lacking a cysteine or having only one
cysteine were f'ound to be useful in assays to detect
metalloproteinase and as means of purifying
metalloproteinases and are also a part of the invention.
Three such structures were peptides having the amino acid
sequences:
DIYGNPIKRIQYEIKQIKKFKGIEKDIEFIYTAPSSAVCGVELDVGGK,
DVGGKKEYLIAGKAEDGKRHITL, and RHITLCDFIVPWDTLSTTQKKSLN.
Peptides of the invention may be used in tests to
assay metalloproteinases in animal or human tissues or in
body fluids which may have antibodies to the protein.
Peptides may also be used to elicit antibodies for use in
detecting metalloproteinases.
The amino acids herein are given the usual one letter
abbreviations accepted as:
A is Alanine C is Cysteine
D is Aspartic Acid E is Glutamic Acid
F is Phenylalanine H is Histidine
I is Isoleucine. K is Lysine
L is Leucine M is Methionine
N is Asparagine P is Proline
Q is Glutamine R is Arginine
S is Serine T is Threonine
V is Valine W is Tryptophan
X is Tyrosine Y is Pyroglutamic Acid
* ~r *
~ '.
,,~.
